Reading the Mask: Descriptor-Grounded BI-RADS Rationale Generation for Breast MRI Segmentation

Key points

• We investigate whether clinically relevant descriptors can instead be derived from the segmentation masks and their dynamic contrast-enhancement series.1
• We have used Breast Imaging Reporting and Data System (BI-RADS) guidelines as a structured framework.1
• Using these guidelines we designed a BI-RADS decision tree with eight quantitative measurements using radiomics, together with the lesion size and component count, that determine the descriptors at eight nodes.1
• After applying this framework to the expert and automatic masks of each of the 1,506 cases in the MAMA-MIA dataset, we generated a reproducible corpus of 3,012 BI-RADS grounded rationales.1

Sources (1)

• [1]Reading the Mask: Descriptor-Grounded BI-RADS Rationale Generation for Breast MRI Segmentation
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 21, 07:50 AM

  Deep learning based models accurately delineate tumours on dynamic contrast-enhanced MRI (DCE-MRI), but segmentation masks alone do not provide the rationales or reasoning used by radiologists to assess the lesion suspicion.

  We investigate whether clinically relevant descriptors can instead be derived from the segmentation masks and their dynamic contrast-enhancement series.

Extractive summary: sentences quoted from the sources.